Determinants of protection against SARS‐CoV‐2 Omicron BA.1 and Delta infections in fully vaccinated outpatients

We aimed to evaluate the association between the humoral and cellular immune responses and symptomatic SARS‐CoV‐2 infection with Delta or Omicron BA.1 variants in fully vaccinated outpatients. Anti‐receptor binding domain (RBD) IgG levels and interferon‐gamma (IFN‐γ) release were evaluated at PCR‐diagnosis of SARS‐CoV‐2 in 636 samples from negative and positive patients during Delta and Omicron BA.1 periods. Median levels of anti‐RBD IgG in positive patients were significantly lower than in negative patients for both variants (p < 0.05). The frequency of Omicron BA.1 infection in patients with anti‐RBD IgG concentrations ≥1000 binding antibody units (BAU)/mL was 51.0% and decreased to 34.4% in patients with concentrations ≥3000 BAU/mL. For Delta infection, the frequency of infection was significantly lower when applying the same anti‐RBD IgG thresholds (13.3% and 5.3% respectively, p < 0.05). In addition, individuals in the hybrid immunity group had a 4.5 times lower risk of Delta infection compared to the homologous vaccination group (aOR = 0.22, 95% CI: [0.05−0.64]. No significant decrease in the risk of Omicron BA.1 infection was observed in the hybrid group compared to the homologous group, but the risk decreased within the hybrid group as anti‐RBD IgG titers increased (aOR = 0.08, 95% CI: [0.01−0.41], p = 0.008). IFN‐γ release post‐SARS‐CoV‐2 peptide stimulation was not different between samples from patients infected (either with Delta or Omicron BA.1 variant) or not (p > 0.05). Our results show that high circulating levels of anti‐RBD IgG and hybrid immunity were independently associated with a lower risk of symptomatic SARS‐CoV‐2 infection in outpatients with differences according to the infecting variant (www.clinicaltrials.gov; ID NCT05060939).

measurement of cell-based immunity implies greater complexity and costs. Therefore, the evaluation of rapid, high-throughput, and easy to perform cell-based techniques such as an interferon-γ release assay (IGRA) as potential CoP for SARS-CoV-2 is of paramount importance to more accurately assess population susceptibility and optimize current vaccine strategies. 11 The objective of this prospective study was to investigate the association of anti-RBD antibody titer and IGRA result with the risk of symptomatic breakthrough infection with Delta or Omicron BA.1 variants in fully vaccinated outpatients.

| Study design and population
This prospective test-negative study was conducted at the University Hospital of Lyon, France (Hospices Civils de Lyon, HCL) between August 23, 2021 and February 2, 2022. The present study was part of a larger study including healthcare workers (HCWs) and non-HCWs with a full vaccination schedule. Clinical, microbiological, and serological data were collected for all included patients at diagnosis.
The inclusion criteria of the larger study were: (i) at least 18 years of age, (ii) written informed consent for participation, and (iii) enrolled in a social security scheme.
For the analysis of the present study, patients were considered if they had (i) a full vaccination schedule, defined as at least two doses of COVID-19 vaccines for infection-naïve individuals or as a previous infection followed by at least one vaccine dose (or vice versa); (ii) presence of COVID-19 symptoms at the time of inclusion, that is, at diagnosis; (iii) maximum interval of 5 days from symptoms onset to diagnosis; (iv) minimum interval of 14 days and maximum interval of 300 days from last immunization (infection or vaccination) and onset of symptoms. Additionally, patients were not considered for the analyses if they had missing information regarding vaccination schemes or regarding previous history of COVID -19. In our study, we grouped patients according to the type of immunization: (i) homologous regimen was defined as one, two, or three doses of the same or different mRNA vaccine in naïve patients (Comirnaty, BNT162b2, BioNTech-Pfizer; and/or Spikevax, mRNA-1273, Moderna), (ii) heterologous regimen as the combination of vaccines with different technologies, that is, viral vector vaccine (Vaxzevria, ChAdOx1-S, AstraZeneca) and an mRNA vaccine (one or two doses) in naïve patients, and (iii) hybrid immunity was defined as an immune response induced by an infection and a vaccination, regardless of which immunization occurred first. All individuals considered COVID-19-naïve were negative for anti-nucleocapsid (N) antibodies.
Of note, a subset of patients (n = 44) infected with Omicron BA.1 were longitudinally monitored for 2 weeks to evaluate the dynamics of Omicron BA.1 infection as well as the immune response following infection and results have been already reported. 12

| Microbiological investigations
Nucleic acid extraction from nasopharyngeal swabs was performed on the automated MGISP-960 workstation using MGI Easy Magnetic Kit (MGI Tech). Quantitative SARS-CoV-2 viral load was determined with qPCR SARS-CoV-2 R-gene kit (bioMérieux) that includes four quantification standards targeting the SARS-CoV-2 N gene: QS1 to QS4 respectively 2.5.10, 6 2.5.10, 5 2.5.10, 4 2.5.10 3 copies/mL of a SARS-CoV-2 DNA standard. NSP were also tested using the CELL control R-GENE (bioMérieux) kit to normalize the viral load per 10 4 cells.
The SARS-CoV-2 variant responsible for infection was confirmed by whole-genome sequencing. The routine SARS-CoV-2 next generation sequencing protocol in our laboratory was based on COVIDSeq-Test TM (Illumina) using Artic V4 or V4.1 primers as they became available. cDNA synthesis and amplification were performed using the COVID-Seq-Test TM (Illumina). Libraries were prepared with the COVIDSeq-Test (Illumina), and samples were sequenced with 100 bp paired-end reads using the NovaSeq 6000 Sequencing system SP flow cell. 13

| Anti-RBD antibodies
The presence of anti-SARS-CoV-2 antibodies was evaluated in 636 patients. Anti-RBD levels were determined using the Siemens Healthineers Atellica ® IM SARS-CoV-2 IgG (sCOVG) kit, a fully automated 2-step sandwich immunoassay using indirect chemiluminescent technology, according to the protocol recommended by the manufacturer's. 14 Antibody levels were reported as binding antibody units (BAU)/mL according to the WHO international standard, 15 using the conversion factor provided by the manufacturer (BAU/mL = U/mL × 21.8).

| Neutralization assays
We selected a subset of samples (n = 138) collected from positive and negative patients during both Delta and Omicron waves.

| IGRA
The cellular response was investigated using an interferon-gamma (IFN-γ) IGRA as previously described 12 using the VIDAS ® COVID stimulation and VIDAS ® 9IFN Research Use Only kits (bioMérieux). As these experiments were performed on fresh whole blood, they were carried out on a subset of 243 included patients. In brief, whole blood was stimulated with a restricted pool of peptides specific to SARS-CoV-2 structural proteins. As well, a negative control without stimulation was performed for each sample. The supernatant was then   Table 1).
There was no significant difference in the interval from last immunization (previous infection or vaccination) to the onset of symptoms between SARS-CoV-2 positive and negative patients in each period; Omicron BA.1-positive patients had a shorter interval since last immunization to the onset of symptoms compared to Deltapositive patients (median 62 days vs. median 152 days, p < 0.0001; Table 1). Overall, a greater proportion of patients had three rather than two immunizations during the Omicron BA.1 period (80.5%) than in the Delta period (10.8%, p < 0.001; Table 2).  Table 1). The frequency of Delta infection was significantly lower (13.3%, 12/90 and 5.3%, 1/19, respectively) when applying the same ≥1000 and ≥3000 BAU/mL thresholds (p < 0.05). In the multivariable logistic regression analysis,

| Impact of immunization regimen on the risk of infection
The majority of patients with homologous immunization regimen (two or three doses) were vaccinated with BNT162b2 during Delta  Table 2).
The median anti-RBD IgG titers of SARS-CoV-2 negative individuals in the hybrid group was significantly higher than in negatives from homologous and heterologous groups during  Table 3.     Table 4). higher than 500 BAU/mL. 22 We estimated a similar level of protection during the Delta period for the homologous (80%; two mRNA doses) and heterologous (89%; one dose of a viral vector vaccine and one dose of an mRNA vaccine) vaccination regimens when we applied the same threshold, or even about three-times higher for the hybrid immunity group (97%). Conflicting results have been reported in the case of Omicron variant regarding the value of anti-RBD IgG as a CoP. [6][7][8]23 The reduction in neutralization titers observed for Omicron compared to previous variants 24  The combination of natural immunity and vaccination is attracting more attention and importance due to the globally increasing SARS-CoV-2 infected and vaccinated population. In the present study, we found higher anti-RBD IgG titers in the individuals with hybrid immunity compared to vaccinated patients regardless of the period, as observed elsewhere. 32,33 Regarding the cellular response, a greater proportion of IGRA-positive patients was observed among those with hybrid immunity in the present study, as described previously for the T cell response using IFNγ ELISpot assays. 34 These findings are in line with those of several studies that suggested that hybrid immunity can confer more effective cross-variant neutralization and long-lasting immunity. [35][36][37] In the present study we did not find a significant difference between the proportion of IGRA-positive and -negative patients nor in the levels of IFN-γ between SARS-CoV-2 positive and negative patients during Omicron  It is important to note some limitations of the present study.

| Neutralization assays
First, many patients were excluded from the analyses to limit heterogeneity of population characteristics, reducing the sample size for some immunization regimens. In this sense, we grouped the patients into three main immunization regimens (homologous, heterologous and hybrid), who mainly had two immunizations during the Delta period and three during Omicron period, but some of them had two, three, or even four immunizations. Although most of the patients were immunized with BNT162b2, it should also be noted that we did not distinguish between the type of mRNA vaccine in the homologous and heterologous regimens (BNT162b2 and mRNA-1273 vaccines). Variability in the number of immunizations may have had an effect on the results and should be evaluated in future studies in a larger number of patients. Similarly, it should be mentioned that the present study was conducted in healthy HCWs with a median age of 36 years, therefore, the results cannot be generalized to older or at-risk populations. However, despite these limitations, as we have used a test negative design, the cases and controls had similar characteristics minimizing confounding by health care-seeking behavior. With regard to neutralization assays, not all samples were tested for the determination of neutralizing capacity since it is a logistically demanding process as it requires live viruses manipulated in a biosafety level-3 laboratory that requires trained staff and specific equipment. Likewise, PRNT 50 was estimated by the evaluation of samples by eye and microscope and thus is highly dependent on the operator, which may also influence the results.
In conclusion, the present study suggests that higher circulating levels of anti-RBD IgG and hybrid immunity were independently associated with a lower risk of symptomatic SARS-CoV-2 infection in outpatients with differences according to the infecting variant.
Anti-RBD IgG serve as correlate of protection but protective thresholds vary depending on VOCs and immunization history, as observed in our study between Delta and Omicron BA.1 variants.